Open a terminal and you get a mostly empty window with a blinking cursor. No buttons, no menus, no hints. That emptiness is the only genuinely hard thing about it.
Here is the whole idea:
you type the name of a program
you press Enter
the program runs and prints its answer
That is it. A terminal is a text box that runs programs.
Why it looks scary and isn't
A graphical app shows you what you can do. Every feature is a button somewhere, so you can find things by poking around.
A terminal shows you nothing. You have to know the name of the thing you want. That is a different problem from “this is advanced” — it is the difference between browsing a menu and ordering from memory. Twenty or so names cover almost everything a beginner does, and you already learned a much bigger vocabulary to use a phone.
Note
Two words get used interchangeably and it is worth untangling them once. The terminal is the window. The is the program running inside it that reads what you type and decides what to do — usually one called bash or zsh. In practice people say “the terminal” for both, and so will this course.
Quick check
Someone says 'run it from the shell'. What are they asking you to do?
Why this matters right now
Every serious AI coding tool is a terminal program. You start Claude Code by typing claude. Codex, Gemini CLI, and most of what comes next work the same way. The agent then does its own work in that same terminal — creating files, running tests, committing code.
Which means two things. First, you need the terminal to start the agent at all. Second, and more importantly, you need to be able to read what the agent is doing. An agent that runs rm -rf build in front of you is either cleaning up or destroying something, and the only way to tell is to understand the command.
Tip
You are not learning the terminal to become a systems administrator. You are learning it so that when an agent shows you a command, you know whether to say yes.
Quick check
An AI agent proposes to run a command you have never seen. What is the useful skill here?
What you are typing into for the rest of this course
The black box beside these lessons is a simulator. It behaves like a real Unix shell for everything this course teaches, and it runs entirely inside your browser tab.
It is an imitation, not a real machine. Nothing you type reaches your computer, your files, or the internet. Commands like git and npm print believable output without actually doing anything. That is deliberate: it means you can run the genuinely dangerous commands here first, get the reflex, and meet them on a real machine already knowing what they do.
Important
The flip side of a safe sandbox: the mistakes it forgives, a real terminal will not. When a lesson says a command is permanent, believe it — the simulator is being polite, your laptop will not be.
Quick check
What happens to your actual computer when you type a command in this course's terminal?
What comes next
The next three lessons get you to the point where the window stops being blank: how to read the line the terminal shows you, how to run your first command, and how to ask the terminal itself what a command does.